Brian Kennemer has commented on Frank Patrick's (or visa versa it seems now) comments on Jack Vinson's comments on the book Getting Things Done and the use of Gantt and PERT charts as part of what project managers do or shouldn't do.
The suggestion that Gantt and PERT charts aren't about "getting thing done," is a bit misleading. If all the PM does is make charts then the statement is probably right. The charts aren't very useful in the absence of "execution." Remember one of the processes of PMBOK "execute the project?"
Being deep, very very deep, in the middle of assembling a 15 to 20 thousand line master schedule, the Gantt and PERT charts are absolutely critical to our success. It's the ONLY way to find the critical path through a 7 year program, identify the risk mitigation "hot points," show that we actually understand the sequence of events for millions of moving parts all coming together on the launch pad and flying into Low Earth Orbit to demonstrate our capabilities to return to human space flight.
Now are Gantts and PERT useful for all projects? I'd say absolutely yes. But in the wrong hands - bad project managers - they can be dangerous; just like Power Point is dangerous in the wrong hands.
More importantly though is that these charts be used for the right reasons:
- Find risk areas through critical path analysis. BTW Microsoft Project has a serious bug on the backward pass of CPM when milestones are included under a summary task. It calculates the slack in the summary task as well as the individual task.
- Show the flow of value through the project "from left to right."
- Show dependencies between team members (IPTs we call them)
- Indicate firm fixed dates and dates that can move to all participants.
Gantt's aren't our favorite format because details can be hidden in the successor and predecessor links. So we make them explicit with intermediate tasks and the Critical Tools PERT Chart Expert tool. Finally we use RISK+ from C/S Solutions to build our three-point assessment of completion dates and establish the probabilistic program margin needed to actually manage the schedule. RISK+ is a very useful tool in the right hands (me), but also dangerous in the wrong hands. The question it answers is "how confident are we that this program can actually be executed?"
As a side bar for anyone who is ready to jump on the evils of probablistic schedule analysis, please hold you horses for an upcoming post. We use it successfully everyday on projects. Most of the "complaints" received on the forums come from those who don't udnerstand the theory, misuse the tools or anrn't in a business domain that can benefit from this type of analytical assessment of a complex network.
So in the end charts are needed to manage projects, but charts are not project management.
